Skip to content

Latest commit

 

History

History
388 lines (301 loc) · 14.6 KB

README.md

File metadata and controls

388 lines (301 loc) · 14.6 KB
Raw

GPUEater Command Line API

Getting Started

GPUEater is a cloud computing service focusing on Machine Learning and Deep Learning. Now, AMD Radeon GPUs and NVIDIA Quadro GPUs are available.

This document is intended to describe how to set up this API and how to control your instances through this API.

Before getting started, register your account on GPUEater. https://www.gpueater.com/

Prerequisites

  1. Create a JSON file in accordance with the following instruction.

At first, open your account page(https://www.gpueater.com/console/account) and copy your access_token. The next, create a JSON file on ~/.eater

{
        "gpueater": {
                "access_token":"[YourAccessToken]",
                "secret_token":"[YourSecretToken]"
        }
}

or

{
        "gpueater": {
                "email":"[YourEmail]",
                "password":"[YourPassword]"
        }
}
  • At this time, permission control for each token is not available. Still in development.

Installing GPUEater API on your system

Install GPUEater API

curl -sL http://install.aieater.com/setup_gpueater_client | bash -
  • We stil don't support API for Windows. Available platforms are Linux and MacOSX.

Run GPUEater API

Before launching an instance, you need to decide product, ssh key, OS image. Get each info with the following APIs.

Get available on-demand product list

This API returns current available on-demand products.

johndoe@local:~$ gpueater products
-----------------------------------------------------
 ---- Images ----
 0: AMD-ROCm1.8.118+TensorFlow1.3 Ubuntu16.04 x64
 1: AMD-ROCm1.8.192+TensorFlow1.8 Ubuntu16.04 x64
 2: CentOS6 x64
 3: CentOS7 x64
 4: NVIDIA-390+CUDA9.0 Ubuntu16.04 x64
 5: Ubuntu16.04 x64
 6: Ubuntu18.04 x64
-----------------------------------------------------
 ---- SSH Keys ----
 0: my_ssh_key
 1: my_sub_key1
-----------------------------------------------------
 ---- Products ----
 0: a1.rx580   InStock(OK) CPU( 2)  MEM(10240)MB  SSD(  80)GB  $0.3458/h  Radeon RX 580 (8G)                      
 1: a1.vega56  InStock(OK) CPU( 3)  MEM(12288)MB  SSD( 100)GB  $0.4794/h  Radeon RX Vega 56 (8GB)                 
 2: a1.vegafe  InStock(OK) CPU( 4)  MEM(15360)MB  SSD( 120)GB  $0.6164/h  Radeon Vega Frontier Edition (16G)      
 3: n1.p400    InStock(OK) CPU( 1)  MEM( 2048)MB  SSD(  20)GB  $0.0992/h  Quadro P400 (2GB)                       
 4: n1.p600    InStock(OK) CPU( 1)  MEM( 2048)MB  SSD(  20)GB  $0.1058/h  Quadro P600 (2GB)                       
 5: n1.p1000   InStock(OK) CPU( 2)  MEM( 4096)MB  SSD(  40)GB  $0.3306/h  Quadro P1000 (4GB)                      
 6: n1.p4000   InStock(OK) CPU( 4)  MEM(10240)MB  SSD(  80)GB  $0.7936/h  Quadro P4000 (8GB)                      
-----------------------------------------------------

Instance launch

Specify product, OS image, and ssh_key for instance launching.


johndoe@local:~$ gpueater launch
-----------------------------------------------------
 ---- Products ----
 0: a1.rx580   InStock(OK) CPU( 2)  MEM(10240)MB  SSD(  80)GB  $0.3458/h  Radeon RX 580 (8G)                      
 1: a1.vega56  InStock(OK) CPU( 3)  MEM(12288)MB  SSD( 100)GB  $0.4794/h  Radeon RX Vega 56 (8GB)                 
 2: a1.vegafe  InStock(OK) CPU( 4)  MEM(15360)MB  SSD( 120)GB  $0.6164/h  Radeon Vega Frontier Edition (16G)      
 3: n1.p400    InStock(OK) CPU( 1)  MEM( 2048)MB  SSD(  20)GB  $0.0992/h  Quadro P400 (2GB)                       
 4: n1.p600    InStock(OK) CPU( 1)  MEM( 2048)MB  SSD(  20)GB  $0.1058/h  Quadro P600 (2GB)                       
 5: n1.p1000   InStock(OK) CPU( 2)  MEM( 4096)MB  SSD(  40)GB  $0.3306/h  Quadro P1000 (4GB)                      
 6: n1.p4000   InStock(OK) CPU( 4)  MEM(10240)MB  SSD(  80)GB  $0.7936/h  Quadro P4000 (8GB)                      

Product > 2

 Selected => "2: a1.vegafe  InStock(OK) CPU( 4)  MEM(15360)MB  SSD( 120)GB  $0.6164/h  Radeon Vega Frontier Edition (16G)"


 -----------------------------------------------------
  ---- Images ----
  0: AMD-ROCm1.9.224+TensorFlow1.10 Ubuntu16.04 x64    
  1: AMD-ROCm1.9.224+TensorFlow1.8 Ubuntu16.04 x64     
  2: CentOS6 x64                                       
  3: CentOS7 x64                                       
  4: NVIDIA-410.48+CUDA9.0 Ubuntu16.04 x64             
  5: Ubuntu16.04 x64                                   
  6: Ubuntu18.04 x64

Image > 0

 Selected => "1: AMD-ROCm1.8.192+TensorFlow1.8 Ubuntu16.04 x64"


-----------------------------------------------------
 ---- SSH Keys ----
 0: my_ssh_key
 1: my_sub_key1

SSH Key > 0

 Selected => "0: my_ssh_key"

Tag > GPUSteak                

Launching...
.
.
.
null

johndoe@local:~$

In the event, the request has succeeded, then the API returns the following empty data. null

In the event, errors occurred during the instance instantiation process, then the API returns details about the error.

Launched instance list

This API returns your launched instance info.

johndoe@local:~$ gpueater instances
-----------------------------------------------------
 ---- Instances ----
 0: Tag(GPUSteak ) a1.vegafe  CPU( 4)  MEM(15360)MB  SSD( 120)GB  $null/h  Radeon Vega Frontier Edition (16G)      
    ssh root@172.105.219.37 -p 22 -i ~/.ssh/my_ssh_key.pem -o ServerAliveInterval=10

johndoe@local:~$

Login

johndoe@local:~$ gpueater login
-----------------------------------------------------
 ---- Instances ----
 0: Tag(          ) a1.vegafe  CPU( 4)  MEM(15360)MB  SSD( 120)GB  $null/h  Radeon Vega Frontier Edition (16G)      
    ssh root@172.105.219.37 -p 22 -i ~/.ssh/my_ssh_key.pem -o ServerAliveInterval=10

Login > 0

root@C-c8cfaf40-80d5-4350-af8e-4901ac52cad5-1:~# ls
based_on_python3_instance  keras_examples     other_samples
check_tensorflow.py        module_checker.py  remote_desktop.txt
deep_learning_yolo_v2      opencl_samples     tensorflow1.3_examples

Terminate instance

Before terminating an instance, get instance info through Launched instance list API. Also, you can directly specify instance_id and machine_resource_id instead of specifing your tag name.


johndoe@local:~$ gpueater terminate
-----------------------------------------------------
 ---- Instances ----
 0: Tag(          ) a1.vegafe  CPU( 4)  MEM(15360)MB  SSD( 120)GB  $null/h  Radeon Vega Frontier Edition (16G)      
    ssh root@172.105.219.37 -p 22 -i ~/.ssh/my_ssh_key.pem -o ServerAliveInterval=10

Terminate > 0

{}

johndoe@local:~$ gpueater instances
-----------------------------------------------------
 ---- Instances ----

johndoe@local:~$

Intaractive command

[Command] [Action] [Args...]

  Example
   > gpueater products


       ___________ssh_key___________
  0 :                       ssh_keys : Listing registered SSH keys.
  1 :               generate_ssh_key : Generating Key Pair.
  2 :               register_ssh_key : Registering an SSH key.
  3 :                 delete_ssh_key : Deleting an SSH key.
       ____________image____________
  4 :                         images : Listing default OS images.
  5 :              registered_images : Listing all user defined OS images.
  6 :                   create_image : Adding an user defined OS image.
  7 :                   delete_image : Deleting an OS image.
      ___________instance___________
  8 :                       products : Listing on-demand products.
  9 :                      instances : Listing launched on-demand instances.
 10 :            change_instance_tag : Change instance tag.
 11 :                         launch : Launch an on-demand instance.
 12 :                      terminate : Terminate an instance.
 13 :                          start : Start an instance.
 14 :                        restart : Restart an instance.
 15 :     emergency_restart_instance : Force restart an instance.
       ___________network___________
 16 :                      port_list : Listing port maps of instance.
 17 :                      open_port : Register port map.
 18 :                     close_port : Delete port map.
 19 :            network_description : Get a network information of instance.
 20 :                     renew_ipv4 : Assign a new IPv4.
 21 :                   refresh_ipv4 : Refresh IPv4 map of instance.
       ___________payment___________
 22 :                       invoices : Listing invoices.
       __________extension__________
 23 :                          login : Login to instance.
 24 :                            get : Get a file from host.
 25 :                            put : Put a file to host.
 26 :                            cmd : Do any command on instance.
 27 :                             ls : File list on remote.
 28 :                           sync : Synchronize files via rsync.
 29 :                         tunnel : Port forwarding local to remote.
 30 :                        jupyter : Start jupyter and port forward.
 31 :                        version : Version of client.
 32 :                           help : Display help.
 33 :                        upgrade : Upgrade API self.

 Action >

You can select an action with interactive navigator. Action type is number or action name. If you have only one instance, instance selector will be skipped.


 Action > login


-----------------------------------------------------
 ---- Instances ----
 0: "GPUSnake  " :    RUNNING :    n1.p400 : CPU 1 : MEM 2048MB : SSD  20GB : "Quadro P400 (2GB)"
    ssh root@107.191.60.187 -p 22 -i ~/.ssh/my_ssh_key.pem -o ServerAliveInterval=10

Welcome to Ubuntu 18.04 LTS (GNU/Linux 4.4.0-124-generic x86_64)

 * Documentation:  https://help.ubuntu.com
 * Management:     https://landscape.canonical.com
 * Support:        https://ubuntu.com/advantage

 * Meltdown, Spectre and Ubuntu: What are the attack vectors,
   how the fixes work, and everything else you need to know
   - https://ubu.one/u2Know

The programs included with the Ubuntu system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.

Ubuntu comes with ABSOLUTELY NO WARRANTY, to the extent permitted by
applicable law.

root@C-b3230350-d9ff-4da9-b4b6-dc70691b1f3d-1:~#

Console API List

Image
Version Action Description
v0.8 images Listing all default OS images
v1.5 registered_images Listing user defined images
v1.5 create_image Adding an user defined OS image
v1.5 delete_image Deleting an user defined OS image
v2.5 import_image Registering an user defined OS image on the internet
v2.0 snapshot_instance Creating a snapshot
v2.0 delete_snapshot Deleting a snapshot
SSH Key
Version Action Description
v0.8 ssh_keys Listing all ssh keys
v1.0 generate_ssh_key Generating Key Pair
v1.0 register_ssh_key Registering an SSH key
v1.0 delete_ssh_key Deleting an SSH key
Instance
Version Action Description
v0.8 products Listing all on-demand instances
v0.8 launch Launch an on-demand instance
v0.8 instances Listing all launched instances
v1.0 change_instance_tag Changing an instance tag
v1.0 start Starting an instance. If the instance is already RUNNING, nothing is going to happen
v1.0 stop[Deprecated] Stopping an instance. If the instance is already STOPPED, nothing is going to happen
v1.0 restart Restarting an instance
v0.8 terminate Terminating an instance
v1.0 emergency_restart_instance Restarting an instance emergently when an instance is hung up
Network
Version Action Description
v1.0 port_list Listing all ports
v1.0 open_port Opening a port for inbound traffic
v1.0 close_port Closing a port for inbound traffic
v1.0 renew_ipv4 Getting a new IPv4 address
v1.2 refresh_ipv4 Refreshing IPv4 map of instance
v1.0 network_description This API reports current network status information
Storage
Version Action Description
v2.0 create_volume Creating an extended volume
v2.0 attach_volume Attaching an extended volume to an instance
v2.0 detach_volume Detaching an extended volume from an instance
v2.0 delete_volume Deleting an extended volume
v2.0 transfer_volume Transfering an extended volume to another region
Subscription
Version Action Description
v2.0 subscription_instance_list Listing all items of subscription instance
v2.0 subscription_storage_list Listing all items of storages volume for subscription instance
v2.0 subscription_network_list Listing all items of subscription networks
v2.0 subscribe_instance Subscribing a subscription instance
v2.0 unsubscribe_instance Canceling a subscription instance
v2.0 subscribe_storage Subscribing a storage volume for subscription instance
v2.0 unsubscribe_storage Canceling a storage volume for subscription instance
v2.0 subscribe_network Subscribing a network product
v2.0 unsubscribe_network Canceling a network product
Special
Version Action Description
v2.5 live_migration Moving a running instance between different physical machines without termination
v2.5 cancel_transaction Canceling a transaction
v2.5 peak_transaction This API reports current status information of a transaction
Payment
Version Action Description
v1.0 invoice_list Listing invoices for on-demand instances
v2.0 subscription_invoice_list Listing invoices for subscription instances
v2.0 make_invoice Obtain a pdf invoice
Extensions
Version Action Description
v1.2 get Copying a file from remote.
v1.2 put Copying a file to remote.
v1.2 cmd Executing an any command on remote.
v1.2 ls Listing files in a remote host
v1.2 tunnel Port forwarding between local and remote.
v1.2 login Logging in a specific instance through the SSH
v1.2 jupyter Starting jupyter on remote with tunneling. This API is available on MacOSX.
v1.0 help Show help.
v1.0 version Display version.
v1.2 upgrade Console API will be upgraded.

License

This project is licensed under the MIT License - see the LICENSE.md file for details